MINING STATISTICS: A SURVEY AND ANALYSIS OF DECISION TREE ALGORITHMS

Abstract

Decision Tree is the most popular and effective approach in knowledge discovery in addition to in records mining. that is used for exploring huge and complicated bodies of statistics so one can find out useful styles. Decision Tree is used as a predictive version which maps observations about an object to conclusions about the item's goal cost. Class algorithm processed an education set containing a fixed of attributes. As the classical set of rules of the Decision Tree ID3, C4.5, C5.0, CART, CHAID, HUNTS algorithms have the deserves of excessive classifying speed, sturdy studying capability and simple production. However, those algorithms also are unsatisfactory in practical application. when it’s used to classify, there does exists the problem of inclining to select attribute that have extra values, and overlooking attributes that have less values. This paper offers awareness on the numerous algorithms of Decision Tree their characteristic, demanding situations, advantage and downside